Skip to content

How to Compress a Video Without Losing Visible Quality

Every compressed video throws information away. The goal is to throw away the information nobody will notice. This guide explains the handful of settings that decide file size, which ones are safe to change, and which ones cost you sharpness.

Updated 15 November 2026 · 4 min read

Why phone videos are so big

Phones record at a high, fixed bitrate so that fast motion never falls apart. A minute of 4K at 60 fps can easily pass 300 MB. Most of those bits are spent on detail that will be thrown away anyway the moment the video is shown on a small screen or uploaded to a social app.

That is why the first compression of a phone video usually saves a lot with little visible change. Later compressions of an already compressed file save less and cost more.

Quality-based (CRF) versus target bitrate

There are two ways to tell an encoder how hard to compress.

  • Constant quality (CRF). You set a quality level and the encoder spends as many bits as each scene needs. Calm shots get few bits, busy shots get more. The size is not known in advance, but the quality is consistent. For H.264, lower CRF numbers mean higher quality: around 18 looks close to the source, around 23 is the common default and around 28 is clearly smaller with some softening.
  • Target bitrate. You set how many bits per second to use, which fixes the file size. It is the right choice when you have a hard limit such as 10 MB for Discord, but quality varies with the content.

Compress Video uses constant quality presets (Smaller, Balanced, Higher). Compress to Size uses a calculated target bitrate.

Compress VideoStart with Balanced. If the result looks soft, try Higher; if it is still too big, lower the resolution rather than the quality.

Resolution and frame rate: the big levers

Halving the resolution in each direction quarters the number of pixels. Dropping from 4K to 1080p is often invisible on a phone screen and makes the file dramatically smaller at the same quality setting.

Watching onResolution that looks sharp
Phone, in a chat app720p
Phone, full screen1080p
Laptop1080p
Large TV or monitor1440p or 4K

Frame rate matters too. A 60 fps clip has twice as many pictures per second as 30 fps. For talking heads, tutorials and screen recordings, 30 fps (or 15 fps for slides) is enough. Keep 60 fps for sport and gaming where smooth motion is the point.

Resize VideoScale to 1080p, 720p or exact pixel dimensions, keeping the aspect ratio.Change Frame RateConvert 60 fps footage to 30 fps or 24 fps without changing its length.

H.264 or H.265

H.264 plays on practically everything made in the last fifteen years. H.265 (HEVC) reaches similar quality at a noticeably smaller size, but it takes longer to encode, especially in a browser, and some older devices, browsers and apps cannot play it.

Use H.264 when you do not know where the file will be played. Use H.265 for your own archive or for recipients you know have recent phones or computers.

Encoding H.265 in a browser is slow because it runs on a single CPU thread. For long videos, start with H.264 and a lower resolution.

A quick checklist

  1. Trim anything you do not need first.
  2. Choose a resolution that matches the screen it will be watched on.
  3. Use 30 fps unless motion really matters.
  4. Start at Balanced quality and look at the result before going lower.
  5. Use 128 kbps audio for music and 96 kbps for speech.
  6. Compress once from the original, not from an earlier compressed copy.

Questions

Does compressing a video always reduce quality?

Yes, technically, because lossy compression discards information. Done well, the loss is invisible at normal viewing size. The first compression of a phone recording usually saves the most for the least visible change.

Why is my compressed video bigger than the original?

The original was already compressed efficiently, or you chose a higher quality or resolution than it had. Try a smaller preset or a lower resolution, or leave the file as it is.

What CRF should I use?

For H.264, 18 to 20 is close to the source, 23 is a good default and 26 to 28 is clearly smaller. The presets in Compress Video map to sensible values so you do not need to choose a number.

Is H.265 always better?

It is more efficient, but slower to encode and not supported everywhere. H.264 is the safer choice when you do not control the device that will play the file.

Can I compress without re-encoding?

Not meaningfully. Removing the audio, trimming or stripping metadata can be done without re-encoding, but making the picture itself smaller always means encoding it again.

Tools for this

More guides